IRC chat logs for #ltsp on irc.freenode.net (webchat)


Channel log from 8 July 2011   (all times are UTC)

00:27TheMatrix3000 has left IRC (TheMatrix3000!~TheMatrix@rrcs-70-61-255-226.central.biz.rr.com, Quit: ~ Trillian Astra - www.trillian.im ~)
00:28vagrantc has left IRC (vagrantc!~vagrant@75-150-46-245-Oregon.hfc.comcastbusiness.net, Quit: leaving)
01:31drdata has left IRC (drdata!~drdata@67.21.176.230, Ping timeout: 255 seconds)
01:47staffencasa has left IRC (staffencasa!~staffenca@128-193-144-253.oregonstate.edu, Ping timeout: 246 seconds)
02:24drdata has joined IRC (drdata!~drdata@67.21.176.230)
02:59drdata has left IRC (drdata!~drdata@67.21.176.230, Quit: drdata)
03:02Parker955 has left IRC (Parker955!~parker@d72-39-201-9.home1.cgocable.net, Ping timeout: 258 seconds)
03:03Parker955 has joined IRC (Parker955!~parker@d72-39-201-9.home1.cgocable.net)
03:11cornel__ has left IRC (cornel__!~cornel@196.192.42.246, Ping timeout: 258 seconds)
03:12cornel__ has joined IRC (cornel__!~cornel@196.192.42.246)
03:36map7 has joined IRC (map7!~map7@ppp118-209-196-175.lns20.mel6.internode.on.net)
03:37redbaritone has joined IRC (redbaritone!~dhall@75.76.184.227)
03:42Parker955 has left IRC (Parker955!~parker@d72-39-201-9.home1.cgocable.net, Ping timeout: 258 seconds)
03:43parker has joined IRC (parker!~parker@d72-39-201-9.home1.cgocable.net)
03:53redbaritone has left IRC (redbaritone!~dhall@75.76.184.227, Quit: redbaritone)
04:52cornel__ has left IRC (cornel__!~cornel@196.192.42.246, Quit: Leaving)
04:52cornel__ has joined IRC (cornel__!~cornel@196.192.42.246)
05:02parker has left IRC (parker!~parker@d72-39-201-9.home1.cgocable.net, Ping timeout: 258 seconds)
05:03Parker955 has joined IRC (Parker955!~parker@d221-66-22.commercial.cgocable.net)
05:12Parker955 has left IRC (Parker955!~parker@d221-66-22.commercial.cgocable.net, Ping timeout: 258 seconds)
05:12Parker955_Away has joined IRC (Parker955_Away!~parker@d72-39-201-9.home1.cgocable.net)
05:21Parker955 has joined IRC (Parker955!~parker@d72-39-201-9.home1.cgocable.net)
05:21Parker955_Away has left IRC (Parker955_Away!~parker@d72-39-201-9.home1.cgocable.net, Ping timeout: 258 seconds)
05:26alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
05:30map7 has left IRC (map7!~map7@ppp118-209-196-175.lns20.mel6.internode.on.net, Ping timeout: 246 seconds)
05:31
<alkisg>
stgraber: I believe the problems you were seeing with nbd compression were caused by nbd-proxy, now that nbd-proxy is disabled by default, is it possible to turn on compression again?
05:32
It makes nbd disk access about 2.5 times faster, it's very significant especially for fat clients
06:11kim_ has left IRC (kim_!ca8a7259@gateway/web/freenode/ip.202.138.114.89, Ping timeout: 252 seconds)
06:12Parker955 is now known as Parker955_Away
06:33alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
06:39jimjimovich has joined IRC (jimjimovich!~jim@77.239.224.238)
06:41tarzeau has left IRC (tarzeau!~gurkan@bee.ethz.ch, Read error: Connection reset by peer)
06:41
<jimjimovich>
Hi Ya'll. Got an LTSP riddle for you. Got 2 thin clients that are completely identical, configured identically, and both running firefox as a local app (connected to an ubuntu 10.04 ltsp server). On one machine, no matter who logs in, Firefox throws certificate and other errors. How can this be?
07:10
<Hyperbyte>
jimjimovich, try running Firefox from a local xterm, see what errors it spits out on the console about this
07:12
<jimjimovich>
Hyperbyte, great idea. Will try :)
07:17alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
07:18
<jimjimovich>
Hyperbyte, no luck. It doesn't throw any errors in the xterm
07:35mis has joined IRC (mis!ca8a7259@gateway/web/freenode/ip.202.138.114.89)
07:36mis is now known as Guest12483
07:36
<Guest12483>
is there any document regarding kiosk on redhat
07:37
#ulteo
07:46Da-Geek has joined IRC (Da-Geek!~Da-Geek@nat/redhat/x-imsiufobwkpdydwj)
07:50nathelm has joined IRC (nathelm!~jon@202-89-181-178.static.dsl.amnet.net.au)
08:02gnunux has joined IRC (gnunux!~emmanuel@194.167.18.244)
08:08dobber has joined IRC (dobber!~dobber@213.169.45.222)
08:13
<Guest12483>
is there any document regarding kiosk on redhat
08:16
<Appiah>
ltsp on redhat? never heard iof that
08:16
well on fedora ...
08:16
<Guest12483>
fedora we can install ....u saying
08:16
<Appiah>
?
08:16
<Guest12483>
any document for fedora
08:16
<Appiah>
I'm just saying
08:16
I never heard of a LTSP install on RedHat
08:17
and the closest thing to RedHat is Fedora
08:17
check topic
08:18
<Guest12483>
Appiah : can we install ltsp on fedora ?
08:18
<Appiah>
sure go ahead..
08:18
"LTSP-5 is the current version that you should be using. Check out Ubuntu, Debian and Fedora for awesome integration."
08:19
<Guest12483>
k , i tried ubuntu 10.04 but kiosk mode is not working
08:19
<Hyperbyte>
jimjimovich, could it have something to do with your network setup? One client being routed through a proxy, another directly onto the internet, maybe?
08:20
jimjimovich, because if Firefox doesn't error, and clients are onto the same server, and it happens regardless of which user.... there must be something else that's different I guess?
08:21
Guest12483, what is your goal exactly? LTSP on Ubuntu or LTSP on Fedora?
08:22
<jimjimovich>
Hyperbyte: That's what I thought too. But all the machines are routed exactly the same, no proxies or anything. It's just really confusing me. I even tried changing the IP address of the client with no effect.
08:22
<Guest12483>
Hyperbyte: i want to intsall ltsp on ubutu ..but kiosk kiosk mode is not working ..so i want to try on other os
08:22
<Hyperbyte>
jimjimovich, you don't run a mac filter somewhere, that prevents strange machines from accessing certain things?
08:23
<jimjimovich>
also, checked lts.conf and dhcpd.conf for any strange things. it's configured just like all the other machines
08:23
<Hyperbyte>
Guest12483, well, then let me save you some time: Ubuntu is far better for LTSP than Fedora.
08:23
<jimjimovich>
Hyperbyte: used to, but now it's turned off
08:23
Hyperbyte: and the machine can access the internet, it just throws errors on https connections for some reason
08:24
<Hyperbyte>
Guest12483, my advice, stick with Ubuntu and wait someone to help you with kiosk mode (I can't, but others will)
08:24
<Guest12483>
k
08:24
<jimjimovich>
Guest12483: I second that advice.
08:24
<alkisg>
10.04 is missing a file (kiosksession) that's why kiosk isn't working, I told viks about it and showed him the bug report but to no avail. :(
08:25
<Guest12483>
alkisg : tell me how to do that on 10.04 if u can send me any link
08:25
<alkisg>
https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/572974
08:25
<Hyperbyte>
jimjimovich, have you tried other browsers? Like, even lynx, perhaps?
08:26
jimjimovich, also, could you show me the completely certificate error? (pastebin)
08:26
*complete
08:28
<alkisg>
Has anyone tried LDAP with the passwd backend? I'm wondering if we could use that for LTSP by default, it would make things a lot easier than ssh screen scraping or libpam_sshauth etc
08:30
http://www.openldap.org/doc/admin24/backends.html#Passwd
08:34
<jimjimovich>
Hyperbyte: Got it! How stupid! The date got reset to 2009 on the thin client BIOS ... so it was rejecting all certs issued after 2009 :)
08:35
<alkisg>
Ah again the time code not working...
08:35
A fix was committed about this
08:40
<Appiah>
alkisg: whats the diff between ldap passwd backend and just using "Normal" ldap auth
08:40
?
08:41
and also kiosk mode in 10.04 is that fixed in other releases and only affects 10.04 ?
08:41
<alkisg>
Appiah: I've never used ldap, but I imagine that the passwd backend allows you to continue working as usual on the server. E.g. gnome users-admin to manage the user accounts.
08:42
So the ldap server would only be used for thin/fat clients, it wouldn't be needed for the server itself
08:42
(but using the same tools for user management is my main point)
08:42
According to the bug report, the kiosk mode was fixed in natty
08:42
<Appiah>
I've used ldap on many setups
08:42
<alkisg>
So it's broken in 10.04, 10.10 - I don't know about 9.10.
08:43
<Appiah>
and the teachers/admins usally use tools like webmin so I never thought about that
08:43
<alkisg>
Not all people are fond of webmin, many hate it in fact
08:43
<Appiah>
but in cli a root user can reset a ldap password
08:43
like it was a localuser
08:43
dunno about the gui tools
08:44
well lucky there are more tools similar to webmin
08:44
not just webmin
08:44
<alkisg>
Yes, but using the usual tools would be even better
08:44
<Appiah>
I dont like it myself but have gotten the request to set it up
08:44
<alkisg>
Do you think it would work? LDAP with the passwd backend?
08:44
<Appiah>
and I end up customizing it for the teacher so the only thing they have is the users list so they can easily reset password
08:44
<alkisg>
It says it's for "demonstration purposes only"...
08:45
<Appiah>
well I dont know enough about LDAP passwd backend
08:46
<alkisg>
Usually, are ldap clients able to change their passwords etc? Or special tools are needed?
08:46
<Appiah>
if cli can change the password on ldap users "passwd anyuser" , shouldn't the gui tools do the same
08:46
nope
08:46
you tell /etc/nsswitch to look for ldap users
08:47
<alkisg>
So if I install LDAP on a school server, and have the clients ask him for authentication, can the users use "gnome-about-me" to change their LDAP passwords?
08:48
In general, I don't think it's possible, isn't that why special tools were developed for LDAP user management?
08:48
<Appiah>
Like I said, works in a terminal
08:48
dunno about gnome utilites
08:48
<alkisg>
OK, thanks. If someone has access to such a setup and could try it, I'd appreciate it.
08:49
<Appiah>
I'll see if I get time later today
08:49
can't promise anything
08:49
<alkisg>
np
08:55
<Hyperbyte>
jimjimovich, glad you got it sorted. :)
09:03alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
09:53Trixboxer has joined IRC (Trixboxer!~Trixboxer@office.supportdepartment.net)
09:54rishi__ has joined IRC (rishi__!~Trixboxer@office.supportdepartment.net)
10:24nathelm has left IRC (nathelm!~jon@202-89-181-178.static.dsl.amnet.net.au, Quit: Leaving)
10:29jimjimovich has left IRC (jimjimovich!~jim@77.239.224.238)
10:43XTaran has left IRC (XTaran!~abe@kiva6.ethz.ch, Read error: Operation timed out)
10:43XTaran has joined IRC (XTaran!~abe@kiva6.ethz.ch)
11:03mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 260 seconds)
11:05mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
11:10alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
11:10mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 252 seconds)
11:17mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
11:19
<stgraber>
alkisg: hey!
11:19
alkisg: I'm switching to nbd-server for Oneiric
11:20
<alkisg>
Hi stgraber
11:20
<stgraber>
alkisg: so it's going to be using tcp/10809 and config in /etc/nbd-server/config
11:20
instead of inetd
11:20
<alkisg>
Coool! At last!
11:20
<stgraber>
(inetd is broken for nbd apparently and I don't want to debug it ;))
11:20
nbdroot should be easy to implement, nbdswap might be a bit trickier
11:20
<alkisg>
Please also put boot=ltsp_nbd in the pxelinux.cfg/default command line
11:21
For nbdswap we can make a single swap file for all clients (sparse) and export it as cow
11:21
This way nbd-server will automatically clean the temp files as well
11:22
<stgraber>
ah yes, that should work
11:22mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Remote host closed the connection)
11:22mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
11:22
<stgraber>
for boot=ltsp_nbd I'm not going to do anything there yet as Oneiric will very likely port initramfs-tools to upstart
11:22
so it's a bit pointless changing that and then having to rewrite everything anyway with the switch to upstart...
11:22
<alkisg>
Ah... I thought for 12.10 Ubuntu was going to drop upstart?!
11:23
And switch to systemd (not sure about the name) that debian and opensuse use?
11:24
<stgraber>
hmm, rumors :) The only thing that was said is "If we were to switch away from upstart, it wouldn't be possible before 12.10"
11:24
<alkisg>
Mhm
11:24
<stgraber>
and I'd be surprised if that was to happen considering the amount of time that's currently being spent on improving upstart
11:26mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 246 seconds)
11:27mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
11:28
<alkisg>
stgraber: about the rc_whitelist thing... if you'll be changing that now, then I think that the reboot/poweroff upstart scripts would no longer be needed, as sendsigs would work properly
11:32mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 258 seconds)
11:32
<stgraber>
hmm, IIRC the problem was that sendsigs would also kill nbd-client and then the whole system would hang
11:33
but that was a while ago, maybe it's nbd aware now :)
11:34mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
11:42cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
11:42mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 252 seconds)
11:48mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
11:50
<stgraber>
wow, that stuff actually works :)
11:50
now, the fun part of updating ltsp-trunk to support it and then figure out what to do in ltsp-update-image
11:53mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 252 seconds)
11:53
<alkisg>
stgraber: the nbd initscripts register themselves with sendsigs so it's working fine. And getting rid of the shutdown/reboot upstart scripts will also allow other services to properly disconnect.
11:54mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:01mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 255 seconds)
12:01mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:10mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 260 seconds)
12:10mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:13
<alkisg>
So now we only need inetd for ldminfod? I don't think it'd be hard to remove that dependency as well... :)
12:14
<stgraber>
yeah, adding the network part directly in the code should be quite easy
12:14
as in, I have 10 lines of python that can do it :)
12:14
<alkisg>
SimpleHttpServer ftw!
12:15mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 255 seconds)
12:16mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:16
<stgraber>
well no, ldminfod isn't a httpserver, but socket + ThreadinMixIn would do the trick in ~10 lines of python
12:16
less if we don't care about the multithreading part
12:18
<alkisg>
I was thinking of http server in case we want to expand it in the future, and it wouldn't make any difference on client side for now.. but sure anything that works
12:21mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 252 seconds)
12:27mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:32mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 246 seconds)
12:38mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:47
<stgraber>
ldm-trunk updated
12:47
now to work on ltsp-trunk, the hard one, ...
12:47mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 264 seconds)
12:48mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
12:50
<Hyperbyte>
!stgraber
12:50
<ltsp>
Hyperbyte: stgraber: rules.
12:50
<Hyperbyte>
Seriously. All you LTSP guys rock. :-)
12:51
I think it's awesome to see active development like this.
12:55mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Remote host closed the connection)
12:55mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
13:00mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 240 seconds)
13:06mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
13:11Da-Geek has left IRC (Da-Geek!~Da-Geek@nat/redhat/x-imsiufobwkpdydwj, Quit: Leaving)
13:13mistik1_ has joined IRC (mistik1_!mistik1@unaffiliated/mistik1)
13:14mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Read error: Connection reset by peer)
13:15mistik1_ is now known as mistik1
13:25mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 246 seconds)
13:25mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
13:30mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 260 seconds)
13:30jammcq has joined IRC (jammcq!~jam@70-91-230-209-BusName-Michigan.hfc.comcastbusiness.net)
13:31nosedrum has joined IRC (nosedrum!~quassel@92.103.117.4)
13:32alkisg1 has joined IRC (alkisg1!~alkisg@ubuntu/member/alkisg)
13:34alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 255 seconds)
13:35mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
13:40mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 252 seconds)
13:46mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
13:53dead_inside has joined IRC (dead_inside!~dead_insi@76.75.3.174)
13:55pedronveloso has joined IRC (pedronveloso!~pedro@static-b4-252-232.telepac.pt)
13:58alkisg1 is now known as alkisg
14:04
<alkisg>
stgraber: do you think now with the nbd named export change, it's a good time to also change the colons that we use (nbdroot=server:host), to the commas-based syntax that the standard nbd uses? https://bugs.launchpad.net/ubuntu/+source/nbd/+bug/594595
14:05
nbdroot=server,port
14:06gnunux has left IRC (gnunux!~emmanuel@194.167.18.244, Read error: Operation timed out)
14:06andygraybeal has left IRC (andygraybeal!~andy.gray@obsidian.casanueva.com, Ping timeout: 260 seconds)
14:07
<alkisg>
Wouter said he preferred commas to allow for ipv6 later on
14:08
<stgraber>
hmm, I find the idea of using commas quite weird... especially as the syntax for ipv6 is quite clear on that
14:09
[2001:470:b511:1000:223:14ff:fea3:d9a4]:10809 is my current nbd server
14:09
<alkisg>
He also allows for a third parameter, server,port,device - but we just use the first one available anyway
14:10
<stgraber>
I don't particularly mind changing the nbdroot syntax as it's not going to be used much anyway :)
14:10
<alkisg>
I was just thinking that it'd be nicer to have a common syntax, that's all, no strong feelings about it
14:10
<stgraber>
most people were just using nbdport=XXX and will now switch to nbdname=XXX
14:11
yeah common syntax is good, it's just that comma separated kind of sucks (as nothing else uses that)
14:11
<host>:<port>/<name> would have been a lot better in my opinion
14:12
<alkisg>
Agreed, if ipv6 uses [] then that sounds better
14:12
<stgraber>
as it'd have been the same as pretty much all the existing formats of URIs in existence :)
14:12
http uses http://[ipv6]:<port>/<page> nfs uses [ipv6]:port/<share>, ...
14:13* alkisg has proposed using dhcp root-path for nbd named exports too
14:13
<alkisg>
But Wouter said he didn't like it because root-path should be used only for nfs (don't know why)
14:14
<stgraber>
option root-path text;
14:14
This option specifies the path-name that contains the client's root disk. The path is formatted as a character string consisting of characters
14:14
from the NVT ASCII character set.
14:14
it's not NFS specific
14:14
and it matches our use of nbd quite well ;)
14:14
<alkisg>
:)
14:15
<stgraber>
anyway, we don't really need wouter's approval to do so. We already have our custom udhcp script for that ;)
14:16
<alkisg>
Of course not, on the contrary, since http and nfs use [] then I'll suggest on the bug report that wouter changes the nbd initscript to match ours :)
14:17
What I'd like though is to be able to boot the same chroot with NFS without running ltsp-update-kernels
14:18
Let me find a tiny change for that...
14:18
ltsp-trunk/client/initramfs/conf.d/ltsp
14:18
BOOT=${BOOT:-ltsp_nbd}
14:19
So if BOOT=nfs is defined on the command line, conf.d/ltsp won't forcifully override it
14:19
Also my experience with compcache so far says that it makes more bad than good for ltsp, I suggest we disabled it by default
14:21
<stgraber>
compcache no longer exists in oneiric anyway, there's something else that's replacing it
14:23* highvoltage never liked compcache
14:23
<stgraber>
the new thing comes directly from the upstream kernel, so maybe it works better
14:25
<alkisg>
The difference between live CDs to ltsp installations, is that in ltsp we have NBD_SWAP, so it's almost always better to use that 25% that compcache/ramzswap reserve as real ram instead
14:32cornel__ has left IRC (cornel__!~cornel@196.192.42.246, Quit: Leaving)
14:35
<highvoltage>
a long time ago in Windows I used to use a tool that did something similar to compcache, but it grew the compressed space as required so it wasn't so intrusive. I hope the new kernel one does something similar
14:36
<alkisg>
Another idea for swap that worked fine in my test was to export the nbd disks as cow, and create a /swap file on them
14:36
This way even aufs wasn't needed
14:37
But nfs won't support that method
14:37gnunux has joined IRC (gnunux!~emmanuel@194.167.18.244)
14:43alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
14:55alexqwesa has joined IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net)
15:12staffencasa has joined IRC (staffencasa!~staffenca@128-193-144-253.oregonstate.edu)
15:18cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 240 seconds)
15:20cyberorg has joined IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg)
15:37gnunux has left IRC (gnunux!~emmanuel@194.167.18.244, Quit: Ex-Chat)
15:41dobber has left IRC (dobber!~dobber@213.169.45.222, Remote host closed the connection)
15:51mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 258 seconds)
15:59ogra_ has left IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de, Excess Flood)
16:00ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de)
16:16alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
16:17shawnp0wers has joined IRC (shawnp0wers!~Adium@linuxjournal/staff/shawnp0wers)
16:23Guest12483 has left IRC (Guest12483!ca8a7259@gateway/web/freenode/ip.202.138.114.89, Ping timeout: 252 seconds)
16:23Parker955_Away is now known as Parker955
16:23ogra_ has left IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de, Read error: Operation timed out)
16:48roasted has joined IRC (roasted!~jason@206.82.22.190)
16:48
<roasted>
Anybody ever have issues with audacity in LTSP?
16:48
Only my first user logged in who uses audacity first can utilize it. they create a tmp folder automatically and they own it. If another user tries to open audacity it tanks, saying no temp location found.
16:49
If I rm -rf the folder, the 2nd user can then run audacity. The 1st user? Nope.
16:49
Any ideas?
16:51
<alkisg>
Which folder? /tmp/audacity-$USER ?
16:52
<roasted>
/tmp/audacity-c
16:52
even if I chmod -R 777 it, no dice.
16:52
<alkisg>
Is audacity installed as a localapp?
16:52
<roasted>
It's as if it's based on the owner of tmp/audacity-c, which seems to be whoever opens it first.
16:52
I didn't do much with this install, so I'd have to say no.
16:52
<alkisg>
It should be /tmp/audacity-roasted (i.e. the username there)
16:52
<roasted>
oh
16:53
hm
16:53
<alkisg>
So it's strange that you have -c there
16:53
distro/version?
16:53
<roasted>
alkisg, I see that now. on my laptop it comes up as audacity-jason
16:53
alkisg, I'm on 10.04 Edubuntu 64 bit
16:53
<alkisg>
And I suppose your $USER env var is set, right?
16:54
<roasted>
um..
16:54
can you elaborate?
16:54
<alkisg>
echo $USER
16:54
(from one of the accounts having the problem)
16:54
<roasted>
k, one second, booting them up
16:54
its always the person who tries to open audacity 2nd
16:54
whoever opens it first = no problems
16:55
the ownership of audacity-c seems to be dictated by who opens it first
16:55
<alkisg>
If all of them want to access /tmp/audacity-c, then yes it makes sense for only the 1st to succeed
16:55
<roasted>
not sure I follow. I DO want all of them to access, yet it's failing.
16:55
<alkisg>
Because normally the first would be /tmp/audacity-roasted, the second /tmp/audacity-alkisg etc
16:55
So no, you don't want them to access the same dir
16:56
You want audacity to use different dirs
16:56
<roasted>
oh, I see
16:56
I wonder if I can just set audacitys preferences to /tmp/audacity-$USER
16:56
<alkisg>
First, see why it doesn't work ok
16:56
Then, find workarounds
16:56
:)
16:56
<roasted>
now wiat a second
16:56
I just booted up
16:56
logged in as local administrator
16:57
and I opened audacity and in tmp I see audacity-administrator
16:57
maybe it was just hating on me for a few moments??
16:57
<alkisg>
Maybe for those users you forced the tmp dir to the same location
16:57
<roasted>
ah, there it is
16:57
<alkisg>
Remove their audacity preferences to revert them to defaults
16:57
<roasted>
audacity-c when I logged in as myself on the thin client
16:58
Ill have to reset the profile in ske
16:58
skel
16:58
<alkisg>
$ grep ^TempDir .audacity-data/audacity.cfg
16:58
TempDir=/tmp/audacity-alkisg
16:58
In that file the temp dir is stored
16:58
Delete that file, change that line, something to make it use the default dirs which contain the username
16:58
Did you include that file to your /etc/skel?
16:59
If yes, that's what broke it
16:59
Don't include paths to your /etc/skel
16:59
<roasted>
I think I may have
16:59
I just did a select all to the home dir
16:59
hidden files included
16:59alexqwesa_ has joined IRC (alexqwesa_!~alex@109.172.15.11)
16:59alexqwesa has left IRC (alexqwesa!~alex@alexo-veto.broker.freenet6.net, Ping timeout: 260 seconds)
16:59
<roasted>
if I'm setting my skel, which should I exclude or include then?
17:01
<alkisg>
I don't use /etc/skel, but if I did, I would manually check all the files I put there to make sure they don't contain paths
17:02
I prefer logon scripts for user dir initialization
17:02
<roasted>
I see
17:02
I prefer using default profiles since that's just what I'm used to with windows.
17:02
Ill just copy everything to skel I did before, but exclude audacity
17:02
and see if that fixes this particular issue
17:02
<alkisg>
On windows they don't work too
17:02
<roasted>
I see in the cfg exactly what you were talkin about
17:03
they work fine on windows xp
17:03
on windows vista and 7, its a joke
17:03
<alkisg>
Nope, you just didn't realize it
17:03ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de)
17:03
<roasted>
perhaps I didn't, but they seem to operate normally for us
17:03
<alkisg>
Absolute paths, registry keys, same SIDs etc make a mess
17:03
<roasted>
the only thing that I really do with default profiles in xp is so they get the same printers and firefox settings
17:03
<alkisg>
E.g. if sometimes a PC wouldn't see the network shares of another PC, how would you imagine that the default profile was to blame?
17:04
But anyways, do what works for you
17:04
<roasted>
potentially same ID's trying to talk to one another?
17:04
well, it's what we've always done. I wouldn't say it's optimal by any means. but its done the job.
17:06
alkisg, you, once again, were dead on
17:06
I must owe you 6 cases of beer by now bro...
17:09
<alkisg>
Haha I'll avoid your place then, you'll get me fat again - I spend days and days at the gym trying to get thinner :D
17:10
<roasted>
hahah, well there's always light beer! :P
17:10
but I hear ya, I've been on the same path as you with the gym and self training. It's hard to keep up on.
17:15
well, alkisg, once again you the man. Thanks for your time and effort. It's appreciated like you cannot believe.
17:15
Gotta get rolling though. More imaging to do!
17:16ogra_ has left IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de, Excess Flood)
17:16ogra_ has joined IRC (ogra_!~ogra@p5098ed03.dip0.t-ipconnect.de)
17:16roasted has left IRC (roasted!~jason@206.82.22.190, Quit: Leaving)
17:23alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
17:25adrianorg has joined IRC (adrianorg!~adrianorg@201.22.231.116)
17:25adrianorg_ has joined IRC (adrianorg_!~adrianorg@201.22.231.116)
17:25nosedrum has left IRC (nosedrum!~quassel@92.103.117.4, Remote host closed the connection)
17:34adrianorg has left IRC (adrianorg!~adrianorg@201.22.231.116, Quit: Fui embora)
17:34adrianorg has joined IRC (adrianorg!~adrianorg@201.22.231.116)
17:35adrianorg_ has left IRC (adrianorg_!~adrianorg@201.22.231.116, Quit: Fui embora)
17:56cyberorg has left IRC (cyberorg!~cyberorg@opensuse/member/Cyberorg, Ping timeout: 258 seconds)
18:20shawnp0wers has left IRC (shawnp0wers!~Adium@linuxjournal/staff/shawnp0wers, Quit: Leaving.)
18:25jfisher has joined IRC (jfisher!~jfisher@jf.stuact.uga.edu)
18:27
<jfisher>
Hi, I'm having a little trouble with the RCFILE variable in lts.conf. I've created a short little bash script, copied it to /opt/ltsp/i386/etc/init.d/myscript.sh, and I've added the line "RCFILE_01=/etc/init.d/myscript.sh" to the computer I'm targeting it to in lts.conf, but the script doesn't appear to be running. Am I missing something?
18:27
Oh and I did run a ltsp-update-image command
18:28
And I know other settings in lts.conf are applying correctly to targeted machines
18:28
<knipwim>
perhaps a stupid question, but is it executable?
18:29
<jfisher>
chmod'd a+x
18:29
hopefully it is something stupid, those are usually easy to fix :)
18:30
And I did test run the script manually, it works as intended
18:30
<jammcq>
is it a bash script?
18:30
<jfisher>
it is
18:30
<jammcq>
what's the first line?
18:31
I start mine with '#!/bin/bash'
18:31
<jfisher>
#!/bin/bash
18:31
<jammcq>
hmm
18:31
when debugging this kind of thing, i'll usually add a line like: env >/tmp/env.out
18:32
near the top of the script
18:32
then on the client, look in /tmp for that file
18:32
to see if it's even trying to run
18:32
<jfisher>
good idea, i'll give that a shot
18:32
<jammcq>
cuz knowing the script is even trying to run is a good first step
18:35
<jfisher>
yeah it did print a list of env variables
18:35
<jammcq>
well... that's good to know. what is the script trying to do?
18:35
<jfisher>
so it looks like it ran
18:35
it tries to set the browser homepage, and then run firefox
18:36
sed -i 's|\("browser.startup.homepage",\) "\(.*\)"|\1 "http://nhl.com"|' $HOME/.mozilla/firefox/*.default/prefs.js
18:36
firefox &
18:36
so two lines
18:36
<jammcq>
is '$HOME' set?
18:36
actually, what would it be set to, if you haven't logged in yet?
18:36
<jfisher>
oh, rc files run prior to login?
18:36
<jammcq>
yeah, they run at boot time
18:36
<jfisher>
(I know just enough linux to be dangerous)
18:37
I see.
18:37
<jammcq>
are you trying to run firefox as a local app?
18:37
<jfisher>
No, I don't mind it being on the server
18:37
<jammcq>
cuz you probably want to put this on the server, in /etc/X11/Xsession.d
18:37
so when you log in, it runs that script
18:38
<jfisher>
well I'd like to target the script to specific user groups
18:38
and modify it to launch different webpages
18:38
based on the client
18:38
<jammcq>
I wonder if there's a provision for running Xession scripts for individual users
18:38
hmm
18:38
<jfisher>
I haven't been able to dig up much in my googling
18:42Gadi has joined IRC (Gadi!~romm@ool-18bbe47a.static.optonline.net)
18:50
<knipwim>
is there any way to search through the irc logs?
18:51
<abeehc>
search google with "your query site:http://www.nubae.com/logs/" maybe
18:57
<knipwim>
abeehc: yes!
18:57
genius
19:05bobby_C has joined IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at)
19:08
<knipwim>
and 5.2.9 released for gentoo
19:20shogunx has left IRC (shogunx!~shogunx@rrcs-67-79-182-228.se.biz.rr.com, Ping timeout: 255 seconds)
19:27jhutchins has joined IRC (jhutchins!~jonathan@64-151-37-66.dyn.everestkc.net)
19:35pedronveloso has left IRC (pedronveloso!~pedro@static-b4-252-232.telepac.pt, Quit: Leaving)
19:44tl1000s77 has left IRC (tl1000s77!~tl1000s77@173-13-114-113-NewEngland.hfc.comcastbusiness.net, Quit: Leaving)
19:52Gadi has left IRC (Gadi!~romm@ool-18bbe47a.static.optonline.net)
20:16shogunx has joined IRC (shogunx!~shogunx@rrcs-67-79-182-228.se.biz.rr.com)
20:27jfisher has left IRC (jfisher!~jfisher@jf.stuact.uga.edu, Remote host closed the connection)
20:29jammcq has left IRC (jammcq!~jam@70-91-230-209-BusName-Michigan.hfc.comcastbusiness.net, Quit: leaving)
20:38vmassuchetto has joined IRC (vmassuchetto!~vmassuche@200-233-169-086.xf-static.ctbcnetsuper.com.br)
20:38
<vmassuchetto>
is there a way to make command line auto login?
20:39
i mean, no ldm, just a command line prompt
20:39roasted has joined IRC (roasted!~jason@206.82.22.190)
20:39
<roasted>
How can I default XFCE to be the default DE of my clients?
20:48
<knipwim>
roasted: for fluxbox i got LDM_SESSION = /usr/bin/fluxbox in lts.conf
20:50
<roasted>
hm, I wonder if xfce would be the same
20:50
or should I say, "xubuntu-session"
20:52bobby_C has left IRC (bobby_C!~bobby@85-124-22-227.teleworker.xdsl-line.inode.at, Ping timeout: 260 seconds)
20:55mgariepy has left IRC (mgariepy!~mgariepy@ubuntu/member/mgariepy, Quit: Leaving)
21:06Parker955 has left IRC (Parker955!~parker@d72-39-201-9.home1.cgocable.net, Ping timeout: 258 seconds)
21:06alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
21:06Parker955 has joined IRC (Parker955!~parker@d72-39-201-9.home1.cgocable.net)
21:07
<vmassuchetto>
no clue how to auto login in tty?
21:10
<knipwim>
i think the shell screen script takes you to a shell
21:10mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
21:11
<knipwim>
it's located on the client in /usr/share/ltsp/screen.d/ (on gentoo)
21:16
<alkisg>
vmassuchetto: do you want to login locally on the client, or to start a session on the server with xterm on it?
21:23
<vmassuchetto>
alkisg, i just need to startx
21:23
alkisg, the rest of the things are fine until now
21:23
alkisg, ive put startx on .bashrc, so i think its cool to just auto login on tty
21:24roasted has left IRC (roasted!~jason@206.82.22.190, Quit: Leaving)
21:27
<vmassuchetto>
knipwim, how these scripts are called?
21:28
knipwim, i mean, how can i use them to get in a shell?
21:35Trixboxer has left IRC (Trixboxer!~Trixboxer@office.supportdepartment.net, Quit: "Achievement is not the end, its the beginning of new journey !!!")
21:45
<alkisg>
vmassuchetto: do you know the difference between "running locally" and "running on the server"?
21:45
Usually, the LTSP clients get connected to the server with ssh, which is managed by LDM, the ltsp display manager
21:46
So there's no easy way to autologin on the server from a tty on the client
21:46
You can autologin with LDM though, but that's after X has started
21:46
<vmassuchetto>
alkisg, so maybe using the ldm_autologin
21:46
alkisg, but i would need this option to be enabled for every client
21:46
<alkisg>
Yes. So, delete that startx command from your bashrc
21:47
<vmassuchetto>
alkisg, done
21:47
<alkisg>
Then put LDM_AUTOLOGIN=True under [Default] in lts.conf, so that it affects all clients,
21:47
and put LDM_USERNAME/LDM_PASSWORD separately for each one of them
21:47
If you don't define those, they default to the client hostname (ltsp123 etc)
21:48
It's strongly suggested that you use different user accounts for each client
21:48
<vmassuchetto>
alkisg, is there a way to globally set LDM_USERNAME/PASSWORD?
21:48
i would like all of them to fall into the same user
21:48
theyll be sharing an application
21:48
<alkisg>
E.g. you can't start firefox from the same user twice
21:48
Also, gnome will complain if you start it for the same user twice
21:48
<vmassuchetto>
its not firefox, its a custom pygtk application
21:49
<alkisg>
How many thin clients do you have?
21:49
<vmassuchetto>
by now just a couple
21:49
<alkisg>
In the final setup?
21:49
<vmassuchetto>
but well increase for 10 or more
21:49
maybe 15
21:49
<alkisg>
OK, why creating 10 users is so bad?
21:50
<vmassuchetto>
not bad at all
21:50
<alkisg>
Then better follow the "normal" way to do things, you'll save yourself a lot of trouble
21:50
<vmassuchetto>
true
21:50
<alkisg>
You'll need one user for each client
21:51
<vmassuchetto>
ok...
21:51
<alkisg>
You can "assign" specific users to specific workstations with mac address,
21:51
<vmassuchetto>
we expect these clients to have the machines changed quite frequently
21:51
<alkisg>
or, if you put specific hostnames for each client from your dhcpd.conf, then you could also do it by hostname
21:51
<vmassuchetto>
so i cant use mac addresses, for example
21:52
the dhcp.conf way seems nice
21:52
<alkisg>
That relies on mac addresses too
21:52
(To assign specific hostnames to clients)
21:52
<vmassuchetto>
alkisg, .... ow... true
21:52
well... thats a problem
21:52
<alkisg>
Another way would be to create as many users as the subnet you're using
21:53
So e.g. for a /24 subnet, you'd create 255 users
21:53
<vmassuchetto>
thats not a problem too
21:53
<alkisg>
And then the clients would user whatever user corresponds to the IP they took
21:53
*use
21:53
<vmassuchetto>
humn... sounds like what ill have to do
21:53
<alkisg>
If you name your users like this:
21:53
<vmassuchetto>
so i just create users like ltspXXX
21:54
<alkisg>
username=ltsp123, pass=ltsp123
21:54
<vmassuchetto>
ok...
21:54
<alkisg>
Then you don't even need LDM_USERNAME/LDM_PASSWORD
21:54
You only need to specify LDM_AUTOLOGIN=True once, globally
21:55
vmassuchetto: what happens with the user /home's?
21:55
Will they be saving their data somewhere?
21:55
<vmassuchetto>
no
21:55
<alkisg>
So it's something like a kiosk mode, with clean settings each time?
21:55
<vmassuchetto>
exactly
21:55
<alkisg>
Then forget what I said so far
21:55
Ah wait
21:55
<vmassuchetto>
hehehe... ok
21:56
<alkisg>
Are your clients good enough to run the app locally?
21:56
CPU/RAM-wise?
21:56
<vmassuchetto>
that will depend
21:56
i dont have how to find out wich machine will connect at a time
21:56
so maybe its nice to think in the worst case
21:57
but the application is really simple
21:57
<alkisg>
E.g. if you have a light pygtk app, and your worst client has 512 RAM, then I assume that it can run it locally
21:57
<vmassuchetto>
something like that
21:58
<alkisg>
OK then do it like a kiosk mode, it'll also save you a lot of bandwidth, and user management, etc etc
21:58
<vmassuchetto>
nice
21:58
<alkisg>
Read about the kiosk plugin - you need something very similar, except that you'd run your app instead of firefox
21:58
No users necessary on the server
21:58
No /home, no persistency
21:58
<vmassuchetto>
ok... im looking for it
21:58
thanks by now
21:59
*searching
22:09mistik1 has left IRC (mistik1!mistik1@unaffiliated/mistik1, Ping timeout: 276 seconds)
22:11mistik1 has joined IRC (mistik1!mistik1@unaffiliated/mistik1)
22:13alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Quit: Leaving.)
22:39vmassuchetto has left IRC (vmassuchetto!~vmassuche@200-233-169-086.xf-static.ctbcnetsuper.com.br, Remote host closed the connection)
23:01Parker955 is now known as Parker955_Away
23:16Parker955_Away is now known as Parker955
23:58dead_inside has left IRC (dead_inside!~dead_insi@76.75.3.174, Quit: Leaving...)